Uploaded October 2026 | Updated November 2026, 4 days ago
Reach out to Farzad to implement AI in your small business: https://farzad.fm/consulting
Master Plan: What Elon Musk is Actually Building book: https://a.co/d/04C1qu8Z
Abundance or Collapse book: https://a.co/d/0cQgFdGH
Farzad Mesbahi (@farzad-fm) — humanity's analyst on disruptive innovations and their 1–20 year implications. https://farzad.fm
Exclusive: https://farzad.fm/fznews/the-insane-physics-behind-light
AI's biggest bottleneck isn't the math. It's moving data. Fetching one number from DRAM takes about 6,400 times the energy of adding two numbers, and when an Nvidia A100 trains an image model, 60 to 84 percent of its working energy goes into moving data. Data centers used about 415 terawatt-hours in 2024, and the IEA expects that to more than double by 2030.
In this video I explain photonic computing and silicon photonics in plain language: why copper wires turn fast signals into heat, how lasers, waveguides, Mach-Zehnder modulators and micro-ring modulators put data on light, how wavelength-division multiplexing puts many colors on one fiber, and why co-packaged optics cuts each connection from about 30 watts to about 9.
Optical interconnect is already shipping, from Broadcom's Davisson co-packaged switch to Nvidia's Spectrum-X Photonics. I also cover Lightmatter, Ayar Labs, Celestial AI and Marvell, the 2025 Nature papers on photonic processors, and the honest limits of optical computing for AI: analog precision, converters, lasers, memory and scale. Light already won the job of moving AI's data. Doing the math with light is the next fight, and it's still in the lab.
CHAPTERS
0:00 Jensen Huang's $1,000 plugs
1:05 Why moving data costs more than the math
1:58 Copper's limit
2:25 What photonic computing is
3:47 Silicon photonics: lasers and waveguides
4:09 Modulators: Mach-Zehnder and micro-rings
5:42 Many colors on one fiber
6:03 Co-packaged optics
6:30 Shipping now: Broadcom and Nvidia
8:24 Lightmatter, Ayar Labs and Celestial AI
9:37 Doing AI's math with light
10:59 Lab results in Nature
11:21 The walls: scale, memory, precision
12:42 Lasers, and where copper still wins
13:06 Why it's still a big deal
Want to sponsor my channel? Email: sponsors@dcsocialmedia.co.uk
Use Descript to edit your videos: https://descript.cello.so/5G6jmxS0qeP
Wrap your Tesla using TESBROS: partners.tesbros.com/FARZADMESBAHI
Get $100 off Matic Robots: maticrobots.refr.cc/active-customers/u/farzadm?s=sp&t=cp
Use my referral link to purchase a Tesla product https://ts.la/farzad69506
Want to grow your YouTube channel? DM David Carbutt
For 10% discount quote ‘Farzad’ https://x.com/DavidCarbutt_
I worked at Tesla starting from 2017 thru 2021. I spent most of my time in the distribution and supply chain organizations in leadership positions.
Before Tesla, I was a Director of Business Intelligence and Pricing at the largest Pet Food & Supply distributor in the US, Phillips Pet Food & Supplies. My wife and I also owned a small business in Bethlehem, PA between 2016 and 2019.
I have been a shareholder of Tesla since 2012 and currently own Tesla stock. Nothing I say constitutes as investment or financial advice.
I have been a shareholder of Lemonade since 2025 and currently own Lemonade stock. Nothing I say constitutes as investment or financial advice.
I currently own SpaceX stock. Nothing I say constitutes as investment or financial advice.
I currently own NVIDIA stock. Nothing I say constitutes as investment or financial advice.
I currently own Meta stock. Nothing I say constitutes as investment or financial advice.
My thoughts are my own and are not representative of everyone who currently works, or has worked at Tesla, Phillips, or any other company or organization I discuss on this channel.
If you like this content I would greatly appreciate your likes and shares!
#artificialintelligence #ai #robotics #future #futuretech #elonmusk #tesla #spacex #nvidia #chatgpt #claude
Reach out to Farzad to implement AI in your small business: https://farzad.fm/consulting
Master Plan: What Elon Musk is Actually Building book: https://a.co/d/04C1qu8Z
Abundance or Collapse book: https://a.co/d/0cQgFdGH
Farzad Mesbahi (@farzad-fm) — humanity's analyst on disruptive innovations and their 1–20 year implications. https://farzad.fm
Exclusive: https://farzad.fm/fznews/the-insane-physics-behind-light
AI's biggest bottleneck isn't the math. It's moving data. Fetching one number from DRAM takes about 6,400 times the energy of adding two numbers, and when an Nvidia A100 trains an image model, 60 to 84 percent of its working energy goes into moving data. Data centers used about 415 terawatt-hours in 2024, and the IEA expects that to more than double by 2030.
In this video I explain photonic computing and silicon photonics in plain language: why copper wires turn fast signals into heat, how lasers, waveguides, Mach-Zehnder modulators and micro-ring modulators put data on light, how wavelength-division multiplexing puts many colors on one fiber, and why co-packaged optics cuts each connection from about 30 watts to about 9.
Optical interconnect is already shipping, from Broadcom's Davisson co-packaged switch to Nvidia's Spectrum-X Photonics. I also cover Lightmatter, Ayar Labs, Celestial AI and Marvell, the 2025 Nature papers on photonic processors, and the honest limits of optical computing for AI: analog precision, converters, lasers, memory and scale. Light already won the job of moving AI's data. Doing the math with light is the next fight, and it's still in the lab.
CHAPTERS
0:00 Jensen Huang's $1,000 plugs
1:05 Why moving data costs more than the math
1:58 Copper's limit
2:25 What photonic computing is
3:47 Silicon photonics: lasers and waveguides
4:09 Modulators: Mach-Zehnder and micro-rings
5:42 Many colors on one fiber
6:03 Co-packaged optics
6:30 Shipping now: Broadcom and Nvidia
8:24 Lightmatter, Ayar Labs and Celestial AI
9:37 Doing AI's math with light
10:59 Lab results in Nature
11:21 The walls: scale, memory, precision
12:42 Lasers, and where copper still wins
13:06 Why it's still a big deal
Want to sponsor my channel? Email: sponsors@dcsocialmedia.co.uk
Use Descript to edit your videos: https://descript.cello.so/5G6jmxS0qeP
Wrap your Tesla using TESBROS: partners.tesbros.com/FARZADMESBAHI
Get $100 off Matic Robots: maticrobots.refr.cc/active-customers/u/farzadm?s=sp&t=cp
Use my referral link to purchase a Tesla product https://ts.la/farzad69506
Want to grow your YouTube channel? DM David Carbutt
For 10% discount quote ‘Farzad’ https://x.com/DavidCarbutt_
I worked at Tesla starting from 2017 thru 2021. I spent most of my time in the distribution and supply chain organizations in leadership positions.
Before Tesla, I was a Director of Business Intelligence and Pricing at the largest Pet Food & Supply distributor in the US, Phillips Pet Food & Supplies. My wife and I also owned a small business in Bethlehem, PA between 2016 and 2019.
I have been a shareholder of Tesla since 2012 and currently own Tesla stock. Nothing I say constitutes as investment or financial advice.
I have been a shareholder of Lemonade since 2025 and currently own Lemonade stock. Nothing I say constitutes as investment or financial advice.
I currently own SpaceX stock. Nothing I say constitutes as investment or financial advice.
I currently own NVIDIA stock. Nothing I say constitutes as investment or financial advice.
I currently own Meta stock. Nothing I say constitutes as investment or financial advice.
My thoughts are my own and are not representative of everyone who currently works, or has worked at Tesla, Phillips, or any other company or organization I discuss on this channel.
If you like this content I would greatly appreciate your likes and shares!
#artificialintelligence #ai #robotics #future #futuretech #elonmusk #tesla #spacex #nvidia #chatgpt #claude










